I have some video captured from TV with occasional dropped audio, and I'm looking for a quick and cheap way to detect it. I use tcextract to pull out the mp3, very cheap and simple. Then I was hoping tcscan would be able to tell me if the file has dropped audio
If I use the "cat foo.mp3" method, the results are different each time: $ cat foo-good.mp3 | tcscan -x mp3 -e 32000,16,2 POS -1 [tcscan] MPEG-1 layer-3 stream. Info: -e 32000,16,2 [tcscan] Found 654 MP3 chunks. Average bitrate is 128.00 kbps (cbr) [tcscan] AVI overhead will be max. 654*(8+16) = 15696 bytes (15k) [tcscan] Estimated time is 23544 ms (00:00:23.544) $ cat foo-good.mp3 | tcscan -x mp3 -e 32000,16,2 POS -1 [tcscan] MPEG-1 layer-3 stream. Info: -e 32000,16,2 [tcscan] Found 1564 MP3 chunks. Average bitrate is 128.00 kbps (cbr) [tcscan] AVI overhead will be max. 1564*(8+16) = 37536 bytes (36k) [tcscan] Estimated time is 56304 ms (00:00:56.304) A straight tcscan -i doesn't detect dropped audio: $ tcscan -x mp3 -i foo-bad.mp3 POS 0 [tcscan] MPEG-1 layer-3 stream. Info: -e 32000,16,2 [tcscan] Found 99861 MP3 chunks. Average bitrate is 128.00 kbps (cbr) [tcscan] AVI overhead will be max. 99861*(8+16) = 2396664 bytes (2340k) [tcscan] Estimated time is 3594996 ms (00:59:54.996) I did find a work-around -- "sox in.mp3 -p -r 10 -c 1 out.ogg" will strip out the silent chunks and reveal the dropped audio. Any other way I could do this? David
